HEC-RAS Reservoir Transport Simulation of Three Reservoirs in the Lower Susquehanna River Basin. Mike Langland and Ed Koerkle

Transcription

1 HEC-RAS Reservoir Transport Simulation of Three Reservoirs in the Lower Susquehanna River Basin Mike Langland and Ed Koerkle

2 Topics Background / Project Objectives Data Selection - Sediment and Geometric Input Data Sediment transport calibration / simulation Model results / issues

3 Background USGS collected bathymetry and cores in 1990, 1993, 1996, 2001, and 2008 Document change in sediment storage capacity and size composition Previous USGS HEC-6 Model (1995) Remaining Capacity Implications - Chesapeake Watershed TMDL - PA/NY reduce more to meet goals

4 Changes in Bathymetry with Time only 10-15% of original volume remains to fill to capacity

5 Sediment inputs have been decreasing (about 2/3 less)

6 Objectives Construct, calibrate, and validate a 1-D sediment model for the entire Reservoir system (~33 miles) GOAL - Simulate the loads in and out, bed-form change, and particle size distribution Product - Produce input boundary condition files for Conowingo Reservoir for USACE 2-D model

7 Definitions 1-D sediment model model either erodes OR deposits, not both on same transect Shear stress (SS) force of water acting on the channel sides and bed (different for each particle size) model can only have one) Critical Shear Stress (CSS) shear stress required to mobilize sediments (model can only input one value) Generally, if SS < CSS, then deposition, if SS = CSS, then equilibrium, if SS > SCC, then degradation (scour)

8 Susquehanna River Reservoirs Model Simulation area (~33 m)

9 HEC-RAS Model 3 main steps 1) Prepare Input data sediment and flow 2) Construct Geometric and Hydraulic framework 3) Calibrate to observed data

10 Input Data Sediment - transport curves or estimated daily sediment loads and core data Flow rating curve or actual daily flow data

11 Sediment Transport Curves *Transport curves yielded low mass, used actual sediment data

12 Bed Material Particle Size (cores) Generally, more sand, less silt and clay as upstream distance increases. (19 locations in Conowingo) DOWN STREAM UPSTREAM

13 Bed Material Grouping (cores) Safe Harbor 4 groups Based on particle size and bed thickness Assigned average shear stress based on USACE Sediment Flume data Holtwood 5 groups Conowingo 5 groups

14 HEC-RAS Model 3 main steps 1) Prepare Input data sediment and flow 2) Construct Geometric and Hydraulic framework 3) Calibrate to observed data

15 Model Geometry Geometric Options : Adapt previous HEC-6 model (USGS, 1995) Convert HEC-2 (FIS) model to RAS sediment model Construct new RAS sediment model

16 Options : Adapt HEC-6 model ( USGS, 1995) Performed poorly, no digital files Convert HEC-2 (FIS) model to RAS sed model Covers 75% of reach, XS stationing errors, no XS bathymetry, and poor alignment current bathymetry Construct new RAS model Alignment of XS cut lines with current bathymetry Model geometry better suited for sediment model (i.e., no structures, fewer XS) Use Lidar-derived topography for channel banks

17 Flood Insurance Study Cross-sections No transect data

18 Options : Adapt HEC-6 model ( USGS, 1995) Performed poorly, no digital files Convert HEC-2 (FIS) model to RAS sed model Covers 75% of reach, XS stationing errors, no XS bathymetry, and poor alignment current bathymetry Construct new RAS model Alignment of XS cut lines with current bathymetry Model geometry better suited for sediment model (i.e., no structures, fewer XS) Use Lidar-derived topography for channel banks

19

20 Final Model Cross-sections 34 in Conowingo 18 In Holtwood 28 in Safe Harbor 80 X-sections Avg one X-section every 0.4 mile SU SQ LSUS Holtwood Dam Safe Harbor Dam Conowingo Dam

21 Hydraulics Hydraulic Options : Discharge Rating Curve Actual daily value discharge Gate and Spillway Simulation

22 Elevation (ft) Gate Representation (Safe Harbor) LSusqReservSed Plan: SedRunInitial 10/3/2012 Safe Harbor Dam Legend EG 01Jan WS 01Jan Ground Ineff Bank Sta Station (ft)

23 Elevation (ft) Reservoir System Hydraulic Representation LSusqReservSed Plan: SedRunInitial 10/19/2012 SUSQ LSUS Safe Harbor Dam Holtwood Dam Legend EG 09Sep WS 09Sep Cri t 09Sep EG 01Jan WS 01Jan Cri t 01Jan Ground Conowingo Dam Mai n Channel Distance (ft)

24 HEC-RAS Model 3 main steps 1) Prepare Input data sediment and flow 2) Construct Geometric and Hydraulic framework 3) Calibrate to observed data

25 Simulation targets ( calibration data) Reasonable estimates of particle size distributions and sediment depth s in the Reservoir System. Bathymetry from 2008 and 2011 surveys Daily streamflow and sediment loads for More detailed sediment load from Sept flood (Tropical storm Lee)

26 Model Calibration Issues Bathymetry data indicates both deposition and scour in same X-Section, 1-D model simulates only one occurrence Modeled fall velocity (silts and clays) about 2X lower (lack of deposition) then expected from literature values and 2-D model, and could not adjust values Model only allows one critical shear stress value, SEDFLUME data indicates wide variability (8x) Increasing the critical shear resulted in an increase in scour (contradictory effect)

27 GSE 2011 XC21

28 Results Model Development Due to uncertainty (fall-velocity and bed sorting), built and verified 2 models, one net depositional one net scour Both boundary condition outputs delivered to Steve (USACE) for 2-D model Depositional model recommended and produced best overall results Scour model performed better for T.S. Lee and other short-term high flow scour events Allows for range in uncertainty

29 ll (tons) HEC-RAS Deposition Model (Transport Function Laursen (Copeland), Sorting Method Exner5, Fall Velocity Method Ruby, Cohesive shear lbs/sqft)) SUSQ-LSUS 31Dec :00: Conowingo Dam Holtwood Dam Safe Harbor Dam 31D T.S. Lee September 7-13, 2011 Mass In 22.3 M tons Mass In 9.95 M tons Mass Out 20.2 M tons Mass Out 9.57 M tons Difference 2.1 M tons deposition Difference <1 M tons deposition , ,000 Distance Upstream (ft)

30 31Dec :00:00 HEC-RAS Deposition Model Bed Elevation Change (ft) 31Dec :00:00 Legend Safe Harbor Dam Holtwood Dam Conowingo Dam

31 t Cum: All (tons) HEC-RAS Scour Model (Transport Function Laursen (Copeland), Sorting Method - Active Layer, Fall Velocity Method Van Rijn, Cohesive shear lbs/sq ft)) SUSQ-LSUS 31Dec :00: Conowingo Dam 31DEC2 Holtwood Dam Safe Harbor Dam T.S. Lee September 7-13, 2011 Mass In 22.3 M tons Mass In 9.9 M tons Mass Out 25.2 Mass Out 11.4 M tons Difference 2.9 M tons scour Difference 1.5 M tons scour , ,000 Distance Upstream (ft)

32 31Dec :00:00 31Dec :00:00 HEC-RAS Scour Model Bed Elevation Change (ft) Safe Harbor Dam Legend Holtwood Dam Conowingo Dam

33 Model Results Sediment Transport (tons) Loads (tons) HEC-RAS (depositional) CY difference TS Lee (Sept 7-13, 2011) difference Marietta IN 22,300,000 9,950,000 Conowingo IN 22,100, ,000 10,100, ,000 Conowingo OUT 20,200,000 2,100,000 9,570, ,000 HEC-RAS (scour) Marietta IN 22,300,000 9,950,000 Conowingo IN 24,400,000-2,100,000 10,300, ,000 Conowingo OUT 25,200, ,000 11,400,000-1,100,000 USGS ESTIMATOR Marietta IN 22,300, ,950, Conowingo OUT 21,100,000 1,200,000 13,500,000-3,500,000 WRTDS (B. Hirsch) Conowingo OUT WY ,500, TS Lee 18,800,000 --

34 Model results Particle Size HEC-RAS (depositional) Particle Size TS Lee Historic Particle Size Sand/Silt/Clay Sand/Silt/Clay Sand/Silt/Clay Marietta IN 10 / 47 / / 47 / 42 9 / 47 / 44 Conowingo IN 3 / 47 / 50 5 / 50 / 45 n/a Conowingo OUT 1 / 32 / 67 2 / 50 / 48 1 / 51 / 48 HEC-RAS (Scour) TS Lee Historic P.S. Sand/Silt/Clay Sand/Silt/Clay Sand/Silt/Clay Marietta IN 10 / 47 / / 47 / 42 9 / 47 / 44 Conowingo IN 2 / 48 / 50 5 / 51 / 44 n/a Conowingo OUT 1 / 45 / 54 2 / 52 / 46 1 / 51 / 48 Minor differences between models Good correspondence with historic particle size

35 Summary HEC-RAS generally not condusive for cohesive (silts/clays) simulations The 2 models provide a range of uncertainty in the boundary condition files Estimated total sediment transport most likely underestimated but reasonable Both models indicate upper 2 reservoirs still play a role in sediment transport